We and others have found a significant association of rare H-ras VNTR alleles with breast cancer, however, the biological mechanism underlying this relationship remains unclear. Two hypotheses can explain this association. The first is that rare H-ras VNTR alleles are caused by a mutation in another critical gene and are therefore markers of genomic instability, but are not directly responsible for the inherited susceptibility to cancer. The second possibility is that rare alleles directly participate in carcinogenesis via their transcriptional enhancer effects. The studies proposed here are aimed at distinguishing between these two alternate hypotheses. Our recent minisatellite variant repeat (MVR) analyses of the H-ras VNTR show that most rare alleles are characterized by aberrant internal sequences suggestive of DNA rearrangements. Using technology developed in the Garrett study population, we will further explore the relationship of H-ras rare alleles to genetic instability, both in the germline and in tumors, in the larger population of the Carolina Breast Cancer Study (CBCS). We will determine the association of germline H-ras rare alleles with the presence of genetic instability in breast tumors, as measured by microsatellite instability and gene amplification, the inter-generational stability of the H-VNTR both in CEPH pedigrees and in relatives of CBCS breast cancer patients with rare alleles, interactions between H-ras rare alleles and potential epidemiologic and environmental factors for breast cancer and modification by rare H-ras alleles of the risk for breast cancer associated with the known cancer susceptibility genes, BRCA1 and MSH2. In order to determine whether the H-ras VNTR plays a direct role in the development of breast cancer, we will evaluate the putative function of this region as a transcriptional enhancer, determine whether the regulatory effects are mediated by NF-kappaB transcription factors, and determine if rare H-ras alleles have abnormal regulatory functions as compared with their common progenitor alleles. We will also determine whether common alleles are preferentially deleted in breast tumors which exhibit loss of heterozygosity since this would support a direct role for this region in the carcinogenic process. Finally, we will expand our repertoire of molecular techniques to more fully characterize the H-ras VNTR for both length and internal structural variation. By increasing our level of detection of 5' sequences and by developing 3' MVR methodologies, we will obtain complete MVR sequence for the entire length of each allele, thus ensuring that all variant/rare alleles are detected, especially those which arose by recombination.